REVIEW Constraints on Legal Norms: Kelsen's View in the Essays* Stanley L. Paulsont The last years of 's life, between the publication in 1967 of an English translation of his great and difficult work, the Pure Theory of ,I and his death in 1973,2 were marked by a flurry of Kelsenite scholarship.3 For Anglo-American readers, the most important new development was the publication in December of 1973 of a major collection of translations of Kelsen's papers, the Hans Kelsen: Essays in Legal and Moral Philosophy.4 This volume contains the first English translations of a number of Kelsen's note- worthy papers on the theory and all of his papers on the 5 logic of legal norms.

* Hans Kelsen: Essays in Legal and Moral Philosophy. 0. WEINBERGER, ed., P. HEATH, transl. Reidel Publishing Co., Boston, Mass. 1974. Pp. xxvii, 300. $36.00. t Assistant Professor of Philosophy, Washington University (St. Louis). I am indebted to a number of people for helpful suggestions and discussion. I especially want to thank Bonnie Paulson and Carl Wellman. I H. KELSEN, REINE RECHTSLEHRE (2d ed. 1960)[hereinafter cited as REiE REcHTSLEHRE]. The English translation (M. Knight transl. 1967) is incomplete and at some points inaccur- ate; quotations in the text are my own translations from the REINE RECHTSLEHRE, and are cited by section number rather than page number to facilitate reference to the English translation. 2 Among the Kelsen memoria see Akzin, Hans Kelsen-In Memoriam, 8 ISRAEL L. REv. 325 (1973); Gross, Hans Kelsen, 67 Am. J. INT'L L. 491 (1973); Verdross, Hans Kelsen in memoriam, 24 OSTERREICHISCHE ZEITSCHRIFT FOR 6FFENTLICHES RECHT 241 (1973). 1 In 1968 there appeared, in German, a monumental collection of the papers of Kelsen, Adolf Merkl, and Alfred Verdross-a tribute to the prodigy of these theorists, the leading figures in the Vienna School of Legal Philosophy. DIE WIENER RECHTSTHEORErISCHE SCHULE (H. Klecatsky, et al., eds. 1968). This impressive compilation, in two volumes, includes forty- five of Kelsen's major papers in legal philosophy. In 1969, Kelsen's long-time bibliographer and friend, Rudolf M6tall, published the first Kelsen biography. R. MErALL, HANS KELSEN: LEBEN UND WERK (1969). The appendices to M~tall's biography include the most recent of his Kelsen bibliographies. In 1971, the year of Kelsen's ninetieth birthday, two new Festschriften appeared: FEsTScHRIFT FOR HANS KELSEN ZUM 90. GEBURTSTAG (A. Merkl, et al., eds. 1971) (which contains a supplementary bibliography compiled by M6tall) and A Tribute to Hans Kelsen, 59 CALIF. L. Rzv. 609-872 (1971). (0. Weinberger ed., P. Heath transl. 1973) [hereinafter cited as ESSAYS]. On natural law, morality, and related themes (with date of first publication): God and the State (1923), The Idea of Natural Law (1928), State-Form and World-Outlook (1933), What is ? (1953), and Value (1960), Law and Morality (1960), and The Founda- tions of the Theory of Natural Law (1963). On the logic of legal norms: Derogation (1962), On the Concept of Norm (1965), Law and Logic (1965), Law and Logic Again. On the Applica- bility of Logical Principles to Legal Norms (1967), and On the PracticalSyllogism (1968). Three of these papers have appeared previously in English. A different English version of What is Justice?, based on Kelsen's 1952 valedictory lecture at the University of California 19751 Constraints on Legal Norms

Kelsen's views on natural law are widely known. Contending that the natural law theorist conflates the amoral "ought" of the positive or enacted law with the moral "ought" of the natural law, Kelsen argued in a number of his writings that the legal "ought" is to be sharply distinguished from the moral "ought". Kelsen's work on the logic of legal norms, on the other hand, is a new undertaking. He did not begin writing papers on questions of legal norms until the 1960s. These investigations on norms, first published in Euro- pean periodicals and Festschriftenbut largely unnoticed in England and America,' are important not only as a late development in Kelsen's thought and as a contribution to the resolution of concep- tual problems in the developing field of deontic logic, 7 but also as an extreme statement of legal . No review of this length could deal systematically with the full import of the Essays. This discussion focuses on the role of con- straints on legal norms, a pervasive theme in Kelsen's earlier work that is raised anew in the Essays. Part I describes the kinds of constraints with which Kelsen is concerned. Part II examines the doctrines of legal validity and the legal proposition, both of which underlie Kelsen's views on constraints. Finally, Parts Inl and IV analyze Kelsen's views on constraints and raise certain difficulties of a logical nature that are generated by Kelsen's defense of these views.

I. CONSTRAINTS ON LEGAL NORMS Constraints that limit the possible "substance" or "content" of legal norms might be termed substantive constraints. A familiar at Berkeley, appeared in H. KELSEN, WHAT IS JUSTICE? 1-24, 376-78 (1957). Derogation (K. Buschmann transl.) is reprinted, without change, from ESSAYS IN HONOR OF 339-55 (R. Newman ed. 1962); there is no published German version. Norm and Value (M. Knight transl.) is reprinted, without change, from 54 CALIF. L. REV. 1624-29 (1966), which, in turn, is a translation of REINE RECHTSLEHRE § 4(e). A shortened English version of a fourth paper, Law and Logic, appeared, with minor textual changes, in PHILOSOPHY AND CHIuSTIAN- rY: PHILOSOPHICAL ESSAYS DEDICATED TO PROFESSOR DR. HERMAN DOOYEWEERD 231-36 (1965). The ESSAYS also includes two papers on causality: The Emergence of the Causal Law from the Principle of Retribution (1939) and Causality and Accounting (1960). Each, in several different translations, has appeared previously in English. ' But see Hart, Kelsen's Doctrine of the Unity of Law, in ETHICS AND SOCIAL JUSTICE 171, 182-90 (H. Kiefer & M. Munitz eds. 1968); Munzer, Validity and Legal Conflicts, 82 YALE L.J. 1140, 1162-68 (1973). Both Hart and Munzer allude to Kelsen's paper on Derogation, ESSAYS at 261-75. 1 See generally the pioneering work of G.H. von Wright, including Deontic Logic, 60 MIND 1 (1951), reprinted in LOGICAL STUDIES 58-74 (1957); NORM AND ACTION (1963); AN ESSAY IN DEONTIC LOGIC AND THE GENERAL THEORY OF AcTION (1968); On the Logic and Ontology of Norms, in PHILOSOPHICAL LOGIC 89-107 (J. Davis, et al., eds. 1969). The University of Chicago Law Review [42:768

example of a substantive constraint is the recent appeal to the doc- trine of privacy to prevent lawmakers from prohibiting the use of 8 artificial contraceptives. Some substantive constraints are merely contingent, that is, they may or may not exist in a legal system, while others appear to be inherent in any legal system. Privacy, for example, has evolved as a doctrine of constitutional stature only in the last decade; it is conceivable, though unlikely, that the doctrine might be abrogated. But abrogation is inconceivable in the case of, say, norms prohibiting killing or other forms of extreme violence to the person. This difference between the privacy doctrine and criminal law norms prohibiting killing marks an important distinction be- tween what might be termed nonsystemic and systemic substantive constraints. Unlike the privacy doctrine, which as a nonsystemic constraint may be peculiar to one jurisdiction, norms prohibiting extreme violence to the person are found in every legal order. Is there a philosophical explanation of the systemic character of the latter? Can it be shown why prohibitions on killing are the rule, why permissions to kill (as in self-defense) are exceptions, and why man- dates to kill (as in a state execution) are limited to altogether ex- traordinary contexts? Natural law theories have explained the systemic character of these constraints in terms of a teleological theory of human nature. The teleological theory develops the thesis that a man's nature im- pels him toward certain goals or ends. In its traditional form, the theory is a part of a more fundamental theory of morals or theology.' In a more modern form, however, the teleological theory rests not on morals or theology but on the truth of general empirical claims to the effect that man exhibits certain regularities in the goals he chooses.10 In both forms of the theory, constraints on the content of legal norms are necessary to achieve these goals or ends-most ob- viously, self-preservation. Kelsen, while recognizing a variety of nonsystemic constraints on the content of legal norms, denies the existence of systemic con- straints on content. The rejection is part of his polemic against the natural law theory. According to the natural law theory, a is valid

Griswold v. Connecticut, 381 U.S. 479 (1965). See, e.g., The Idea of NaturalLaw, ESSAYS at 27-60; and The Foundationsof the Theory of Natural Law, id. at 114-53, for interpretations of substantive constraints as implications of the natural law theorist's theology. See also text at notes 50-60 infra. 11See, e.g., H.L.A. HART, THE CONCEPT OF LAw 189-95 (1961). 19751 Constraints on Legal Norms

because its content conforms to the natural law qua norm of absolute justice. According to the Pure Theory of Law as a positivistic legal theory, the validity of positive law is alto- gether independent of its content; a positive law is valid not because it has a definite, namely a just content, but because

it was created in a particular way...... But if Kelsen's argument for rejecting the systemic substantive constraints of the natural law theory proves anything, it proves too much. By severing the link between natural law precept and indi- vidual norm Kelsen unwittingly severs the link between authorizing norm and individual norm in his own theory as well.1 2 Kelsen's attack on natural law raises a second issue. If a posi- tive law is valid solely because it has been authorized, then two conflicting norms may exist simultaneously. If, for example, norm A commands Jones to do act x at time t and norm B prohibits Jones from doing act x at time t, then although the norms have the same material elements (subject, act, and time), they are of distinct " modalities"; the first is a command, the second a pro- hibition.13 They present, in virtue of their form alone, an express conflict, 4 for the subject's noncompliance with one of the norms is unavoidable. The problem of expressly conflicting norms might be resolved by invoking a second type of constraint, the constraint on form. While substantive constraints limit the possible content of legal norms, constraints on form limit the possible combinations of legal norms. Like substantive constraints, constraints on form may be in

" Kelsen, Voin Geltungsgrund des Rechts [On the Basis of Legal Validity], in VOLKERRECHT UND RECHTLICHES WELTBILD: FESTSCHRIFT FOR ALFRED VERDROSS 157, 165 (F. von der Heydte, et al., eds. 1960), reprinted in 2 DiE IENER RECHTSTHEORErISCHE SCHULE 1417, 1427 (H. Klecatsky, et al., eds. 1968)(my translation). " See text at notes 59-60 infra. ,3Normative modalities include what Kelsen speaks of as "norms" and also as "norma- tive functions," namely, "to command" (gebieten), "to permit" (erlauben), and "to author- ize" (ermlchtigen). See REINE REcHTSLEHRE §§ 4(b), (d). For the sake of convenience the prohibition is included here as a normative modality although strictly speaking it may be defined in terms of the command and the notion of forbearance; see, e.g., G.H. voN WRIOrHT, NORM AND ACTION ch. 5 (1963). Kelsen introduces another norm or normative function, namely, "to derogate" (derogieren), in his paper on Derogation,EssAYs at 261. The derogating norm, which is not considered here, raises special problems, for it represents a significant departure from Kelsen's conception of the legal norm as developed through the commanding, permitting, and authorizing norms. 11The least problematic case of an express conflict between norms, that between a command and a prohibition, is used here to illustrate the role of systemic constraints on form. For discussion of other, more problematic conflicts between norms, see Derogation, ESSAYS at 269-71; Munzer, Validity and Legal Conflicts, 82 YALE L. J. 1140, 1142-48 (1973). The University of Chicago Law Review [42:768

principle either nonsystemic and ad hoc or systemic and inherent in any legal system. The rule lex posteriorderogat priori is a familiar nonsystemic constraint on form, which resolves conflicts by deny- ing validity to the norm first enacted. The rule is either expressly enacted like other positive legal norms or simply followed as a mat- ter of custom. But it need not be invoked on a given occasion and in fact may not be recognized in the jurisdiction at all. It may be asked whether constraints on form, like substantive constraints concerning, for example, extreme violence, are in fact found in every legal order. Can express conflicts be resolved in terms of systemic features of the norms themselves? For example, might the principle of noncontradiction, which applies to a conjunction of statements that affirm and deny the same proposition, also provide a systemic basis for resolving the problem of expressly conflicting norms? In the Pure Theory of Law (2d ed. 1960), Kelsen invokes the principle of noncontradiction to make a case for one type of systemic constraint on form, although he has abandoned this position in his recent papers on the logic of legal norms in the Essays. The shift is noteworthy as a development of , and it invites at- tention to both the earlier and the later of Kelsen's positions. Al- though the arguments for Kelsen's earlier position are unsound, as he himself recognized in the Essays,'5 in developing these arguments Kelsen has provided a good part of the conceptual rigging necessary for a defensible theory of legal validity-a theory that would pro- vide, inter alia, a structure for both systemic and nonsystemic constraints. Furthermore, if it is true that and laymen alike subscribe to some ill-articulated version of Austinian legal positiv- ism, 6 and if, as I believe, Kelsen's Pure Theory of Law is the most elegant theoretical statement of such views, then an examination of Kelsen's theory may speak to some of our own intuitions about law.

II. FUNDAMENTAL INTUITIVE NOTIONS OF LEGAL POSITIVISM Legal positivism takes as its point of departure the idea that law is conventional, "a human creation, [which] as such presents itself purely as the work of man." 7 Natural law theory, on the other

" See text at notes 65-68 infra. " , among others, suggests that as a conceptual picture of law, the legal positivism of John Austin is "accepted in one form or another by most working and academic lawyers who hold views on ." Dworkin, The Model of Rules, 35 U. CH. L. REV. 14, 17 (1967), reprinted under the title Is Law a System of Rules?, in EssAYs IN LEGAL PHILOSoPHY 25, 28 (R. Summers ed. 1968). "7The Idea of Natural Law, ESSAYS at 30. 1975] Constraints on Legal Norms hand, holds that law "does not rest on a human and therefore inade- quate will . ..but comes about 'on its own', so to speak."' 8 As theories of the nature of law, legal positivism and natural law ex- haust the field-the one, in the words of Lord Bryce, perceiving law as "Artificial, Transitory, and Local" and the other perceiving law as "Natural, Permanent, and Universal."19 One of the fundamental intuitive notions of legal positivism is human will. "From the standpoint of a moral or legal positivism, account is taken only of positive norms, i.e., those posited by real acts of will, and, so far as only man can have a will, of norms posited only by actual human acts of will."2 The second fundamental notion is coercion. Kelsen explains that are "posited by human acts of will" and have, therefore, "an arbitrary character. Any behaviour we please, that is, can be decreed in them to be obligatory."'" But if positive laws are arbitrary and may therefore be unjust, nothing inherent in positive law as- sures compliance. For this reason "coercion [is] an indispensable '22 constituent of positive law. Kelsen develops the fundamental notions of coercion and human will in two different domains-the positive legal order (die positive Rechtsordnung) and legal science (Rechtswissenschaft). In the positive legal order, coercion is manifested in the legal sanction, and acts of human will "create" or "posit" legal norms.2 3 In legal science, the role of the sanction in the positive legal order is charac- terized by the legal proposition, and the necessary conditions for legal norms are set out in the doctrine of legal validity. A. The Legal Proposition The systematic development of the intuitive notion of coercion presents a problem for Kelsen. Coercion is manifested only in the sanction, and the legal norm obscures the role of coercion in the positive legal order. The legal norm in its most familiar form is a command, for example: Jones must do act x at time t. But the command itself does not express a sanction. To fill the gap, Kelsen formulates what he terms the legal proposition

Id. at 28-29. " 2 J. BRYCE, STUDIES IN HISTORY AND JURISPRUDENCE 565 (1901). On the Concept of Norm, ESSAYS at 218. 21 Id. 21 The Idea of Natural Law, ESSAYS at 32. ',Kelsen, Vom Geltungsgrund des Rechts, note 11 supra, at 158. The University of Chicago Law Review [42:768

(Rechtssatz).24 For example, the legal proposition corresponding to the aforementioned norm would read: If Jones fails to do x at t, then 0, a legal official, ought to impose on Jones a certain legal sanction. The legal proposition is hypothetical in form; satisfaction of the condition specified in the protasis ("if" clause) warrants the legal result specified in the apodosis ("then" clause). (The legal result, strictly speaking, consists in "imputing" [zurechnen] liability to the subject, that is, rendering the subject liable to the sanction where he was formerly immune.)2 Some of the differences between legal norms and legal proposi- tions are clear. Norms speak directly to the obligations of legal subjects, prescribing a course of behavior; legal propositions are concerned with the situation in which a subject fails to follow these prescriptions. Norms are created by legal officials (and by private individuals through consensual arrangements), while legal proposi- tions are the constructions of legal theorists. Norms are valid or invalid; legal propositions are true or false. But the nature of the relationship between legal norms and legal propositions is problematic. In his voluminous writings, Kel- sen addresses the question with a bewildering variety of statements. In the Pure Theory of Law (2d ed. 1960), he speaks of the legal proposition as "describing the norm, '26 and earlier, in a paper, as "describing the 'ought' of the legal norm. ' 2 In the General Theory of Law and State (1945), he says that the legal proposition "repre- sents" the legal norm and, in metaphorical language familiar from Kant's doctrine of analyticity, that the legal norm is "contained in" the legal proposition. 2 These cryptic statements are unhelpful, and the question of the nature of the relationship between legal norm and legal proposition persists.

11The translator of the ESSAYS renders Kelsen's term "Rechtssatz" as "legal statement," whereas it is rendered here as "legal proposition," thus facilitating the distinction drawn between Rechtssatze (legal propositions) and Aussagen (statements); see text at notes 64-65 infra. Both translations reflect the fact that for Kelsen, Rechtssatze are descriptive, not normative; earlier translations of Kelsen's "Rechtssatz" as "rule of law" obliterated this critical distinction. See Kelsen's own discussion of the meaning of "Rechtssatz" in Kelsen, ProfessorStone and the Pure Theory of Law, 17 STAN. L. REV. 1128, 1132-37 (1965). See also Woozley, Legal Duties, Offences, and Sanctions, 77 MIND 461, 461-62 n.1 (1968). 25 See REINE RECHTSLEHRE §§ 16-18. See also Moore, Kelsen's Puzzling "Descriptive Ought", 20 U.C.L.A.L. REV. 1269 (1973). 26 REINE RECHTSLEHRE § 16. 2 Kelsen, The Pure Theory of Law and ,55 HARv. L. REV. 44, 51 (1941). 2 H. KELSEN, GENERAL THEORY OF LAW AND STATE 45, 61 (A. Wedberg transl. 1945). 19751 Constraints on Legal Norms

Kelsen addresses the question more precisely in a paper in- cluded in the Essays on "The Idea of Natural Law." Speaking of a norm of the form, "I ought, or am obligated, not to steal," Kelsen says that it "means [bedeuten] in positive law nothing else but that if I steal, I ought to be punished, if I fail to repay a loan received, there should be execution against me. '29 Similarly, albeit in briefer compass, Kelsen remarks in the Pure Theory of Law (2d ed. 1960) that to say "[t]hat behavior is commanded means [bedeuten] that the opposite course of behavior is the condition of a sanction that 'ought' to be imposed."3 Now Kelsen is not suggest- ing here that the meaning of the "ought" of the legal norm is cap- tured by the "ought" in the legal proposition. For the former stands for the will of the law-creating agent and can be represented var- iously by a command, a permission, or an authorization. The "ought" of the legal norm is thus a generic term. 3' By contrast, the "ought" of the legal proposition means simply that where the condi- tion specified by the "if"-clause is satisfied, liability to the sanction of the "then"-clause is "imputed" to the individual. 2 Rather, Kelsen is saying that the norm is equivalent to the legal proposition. The equivalence of legal norm and legal proposition shall be understood, here, to mean that the legal norm is valid if, and only if, the corresponding legal proposition is true.33 For exam- ple, The command "Subject S must do act x at time t" is valid if, and only if, the legal proposition "If S fails to do x at t, then 0, a legal official, ought to impose on S a certain legal sanc- tion" is true. The view that the norm and the legal proposition are equivalent has important consequences with respect to Kelsen's arguments on systemic constraints on form.

B. The Doctrine of Legal Validity Kelsen discusses human will, the other of the fundamental in-

2,The Idea of Natural Law, ESSAYS at 32-33. 10REINE RECHTSLEHRE § 5(a). 11See text at notes 36-38 infra. 32 See note 25 supra. 1 See text at notes 64-65 infra. The notion of equivalence applies directly to "indepen- dent norms," that is, commands (and prohibitions, see note 13 supra), to which the discussion in Part IV is confined. The notion does not, however, apply directly to "dependent norms," that is, permissions and authorizations; their relation to legal propositions is complex and cannot be understood apart from the command on which the permission or authorization depends. See REINE RECHTSLEHRE § 6(e). The University of Chicago Law Review [42:768 tuitive notions of legal positivism, as it is expressed in the acts of individuals, and specifically, in those acts intended to affect the behavior of others. Because the act of an individual is ascertainable in both space and time and is governed by causal laws, Kelsen speaks of its existence as "the existence of a natural fact."3 His concern, however, is not with the act qua fact, but with its meaning: If through some act an individual expresses his intention that another individual is to behave in a particular way . . . the meaning of his act cannot be accounted for by saying that the other individual will behave in the prescribed way, but only by 3 saying that he ought [solien] so to behave. 1 The "ought" in "ought so to behave" is, in Kelsen's view, the first step in an account of the meaning of the individual's act. But "ought" (sollen) has a Pickwickian use in the Pure Theory. 6 The conventional vocabulary of ethics confines the meaning of the verbal auxiliary "ought" to the single normative modality of command. Kelsen, however, uses "ought" as a generic term ranging over the specific normative operators ("must," "may," and "can") of the specific normative modalities (command, permission, and authori- zation, respectively) .7 Corresponding to the generic "ought" is the substantive expression "prescription" (and also "to prescribe"), which ranges over the substantives of the specific normative modali- ties.3 1 In some cases the meaning of the generic "ought" is, in Kelsen's language, merely "subjective"-the intention of a party issuing a prescription. In other cases the generic "ought" also has, for Kelsen, an "objective" meaning independent of the prescribing party's in- tention.39 The difference between subjective and objective meaning is illustrated by the contrast between "the command of a gangster to give [someone] a certain sum of money and the command of a tax officer which has the same subjective meaning as the former."4 Although these prescriptions have the same subjective meaning, the

31 Kelsen, On the Basic Norm, 47 CALIF. L. REV. 107 (1959). REINE RECHTSLEHRE § 4(b). " See id. at § 4(b)-(c). See note 13 supra. 's This discussion follows Kelsen's practice of using the generic "ought" (as well as "prescription" and "to prescribe") to talk indifferently about the specific normative modali- ties. See, e.g., On the Concept of Norm, ESSAYS at 216, where, after alluding to the specific normative modalities, Kelsen speaks generically of norms "as prescriptions [Vorschriften] for the mutual behaviour of men." 19See RErNE RECHTSLEHRE § 4(b). 10 Kelsen, On the Pure Theory of Law, 1 ISRAEL L. REv. 1, 6 (1966). 1975] Constraints on Legal Norms command of the tax officer has, in addition, an objective meaning. Kelsen speaks of the objective meaning of a generic "ought," or prescription, as a norm, "a specific sense-content whose verbal ex- pression is an ought-statement."' And the validity of a norm is 42 simply its existence in a normative order. But it is difficult to distinguish legal norms from prescriptions that have only subjective meaning. Kelsen sees the resolution of this problem as an important task for the Pure Theory. 3 His solution is the doctrine of legal validity, which provides conditions for deter- mining the validity of legal norms. Although these conditions are to be found, in the first instance, in the positive legal order, their complete statement in the doctrine of legal validity includes two analytical constructions of legal science: first, the concept of author- ization in a normative hierarchy and, second, the concept of the basic norm. 1. Authorization in a Normative Hierarchy. The central idea in the doctrine of legal validity is authorization. The authorizing norm confers power on a particular person or determinate body to create law and specifies the conditions for exercising that power.4 Thus, a given norm is identified as legally valid by appeal to its authorizing norm. The authorizing norm, itself a positive legal norm, is similarly identified as legally valid by appeal to its author- izing norm. One determines legal validity in this manner whenever positive legal norms exist on immediately adjacent levels in the normative hierarchy. Suppose, for example, that the legal validity of a criminal sent- ence, an execution by hanging, is in question. 5 The act may be murder or the terminus of a legal proceeding. It is the latter, Kelsen argues, only if the execution is authorized by an individual legal norm, namely a court order. The court order, in turn, is legally valid only if it is within the scope of the general power-conferring norm that authorizes such orders by the court. And the general power- conferring norm is, in turn, legally valid only if it has been created by a legislative body authorized by the constitution to enact general power-conferring norms. 2. The Basic Norm. The identification of legal norms by ap-

" The Foundationsof the Theory of Natural Law, EssAYs at 114.

42 See REINE RECHTSLEHRE § 4(c). See id. at §§ 4(b)-(c), 6(c), 34(a)-(d); Kelsen, ProfessorStone and the Pure Theory of Law, 17 STAN. L. REV. 1128, 1144 (1965). 1 See REINE RECHTSLERRE § 34(c). 45 Id. The University of Chicago Law Review [42:768

peal to their authorizing norms cannot, however, go on intermina- bly. For at some point one arrives at the "historically first constitu- tion" beyond which there are by hypothesis no further authorizing norms in the positive legal order. At this point Kelsen shifts ground, moving from the enacted norms of the positive order to the presup- posed basic norm, a construction of legal science. Kelsen states: If one. . . inquires into the basis of the validity of a historically first constitution . . .and foregoes any answer in terms of a meta-legal authority, such as God or nature, then the answer can only be that the validity of this constitution, the assump- tion that it is a binding norm, must be presupposed.46 Although there is no consensus on what Kelsen means by the basic norm, the Kantian influence is clear. Kelsen says repeatedly that the basic norm is, in the Kantian sense, a transcendental concept. Just as Kant asks how, apart from metaphysics, it is possible to interpret the synthetic a priori propositions of natural science, "so likewise, the Pure Theory of Law asks: How, apart from a metalegal authority such as God or nature, is it possible to interpret the subjective sense of certain acts as a system of objectively valid 4 7 norms?" The concept of authorization in a normative hierarchy and the concept of the basic norm lie at the core of Kelsen's doctrine of legal validity. There is, however, an additional condition of legal validity, that of efficacy. Kelsen defines an efficacious norm as one that is either complied with by the subject or, in the absence of compliance, is "applied" (anwenden) by a legal official's imposition of the sanction." An occasional lapse in the efficacy of a norm does not undermine the validity of the norm, but a complete- loss of efficacy does. Legal norms, Kelsen contends, "are no longer consid- ered valid if they cease to be efficacious." 9 Kelsen thus sees the doctrine of legal validity as an analytical doctrine of legal science, a complete statement of the conditions for the legal validity of norms. To summarize:

46 Id. 4' Id. at § 34(d). 1' Id. at § 4(c). 11Id. at § 34(g). The requirement expressed here, that the efficacy of the individual legal norm is a necessary condition of the validity of that norm, is stronger than that which Kelsen espouses in the earlier GENERAL THEORY OF LAw AND STATE (A. Wedberg transl. 1945). There he holds only that the efficacy of the normative legal order is a necessary condition of the validity of individual legal norms within that order. See id.at 41-42, 118-20, 122. 1975] Constraints on Legal Norms

A norm is legally valid if and only if (1) it has been created in accordance with the conditions specified by its authorizing norm, and the authorizing norm has, in turn, been created in accordance with the conditions specified by its authorizing norm, and so on, for each level of authorizing norms in the hierarchy; and (2) the norm is generally efficacious. The further condition that the basic norm be presupposed is not included in this statement of the conditions of validity, because the presupposition is implicit whenever a norm is identified by appeal to its authorizing norm. Kelsen's doctrine of legal validity is, in the end, a doctrine of the creation of norms. In sharp contrast, the doctrine of legal valid- ity in the natural law theory rests on an appeal to systemic substan- tive constraints, which cannot be created or abrogated in an ad hoc manner but rather are inherent in any legal system.

III. SYSTEMIC SUBSTANTIVE CONSTRAINTS AND LEGAL NORMS Kelsen prepares the way for his rejection of the putative exist- ence of systemic substantive constraints by distinguishing between two models for a normative order, the dynamic and the static." To determine the legal validity of a norm N in a dynamic normative order, one looks to the authorizing norm that confers the power in accordance with which N was created, and then, to the norm that confers the power in accordance with which the authorizing norm was created, and so on. Details of the dynamic model are familiar from Kelsen's doctrine of legal validity." To determine the validity of N in a static normative order, one must ascertain whether N is deducible from higher order norms and, ultimately, from the pre- supposed basic norm. All of the norms of a static normative order "are already contained in the content of the presupposed norm, and they can therefore be deduced from it by way of a logical operation, 5' 2 that of a conclusion from the general to the specific. According to Kelsen, the normative order of natural law follows the static model. He supposes that by rejecting the static model he has rejected, as well, the systemic substantive constraints of the 53 natural law theory. In a statement reflecting his moral relativism, Kelsen contends that "there cannot be any self-evident norms of

See REINE RECHTSLEHRE § 34(b). 5, See text at notes 43-49 supra. 52 REINE RECHTSLEHRE § 34(b). See, on Kelsen's moral relativism, What is Justice?, EssAYs at 1-26. The University of Chicago Law Review [42:768 human behavior" from which legal norms could be deduced. 4 In his paper on "The Idea of Natural Law," however, Kelsen offers a more instructive explanation for his rejection of the systemic substantive constraints of the natural law theory. Here Kelsen argues that even if the content of certain norms in a static system were self-evident and even if other norms could be deduced therefrom, nevertheless deducibility would be irrelevant to the validity of norms. It follows, Kelsen would have us believe, that self-evident norms will not serve as systemic substantive constraints on norms of the positive or en- acted law. To illustrate Kelsen's argument it is useful to look at a natural law theorist's example of deducibility within a normative order. In the Summa Theologica, St. argues that there are two ways of deriving a command from a natural law precept. The first of these is deduction (demonstratio): • . . commands can be traced to natural law in two ways; one, drawn deductively like conclusions from premises . . . [and this] process is like that of the sciences where inferences are demonstratively drawn from principles .... To apply [this first process] some commands are drawn like conclusions from natural law, for instance, "you must not commit murder" can be inferred from "you must do harm to nobody."5 Aquinas' example of the derivation of a command from a precept of the natural law by deduction is, strictly speaking, enthymematic. That is, if the example represents a logically correct argument, it is a truncated one, for a premise has been omitted. But the omission is innocuous, if the premise in question, "Murder is a form of harm," may be understood as a definition, contributing nothing of substantive import. The addition of the omitted definitional prem- ise then yields the following logically correct argument: You must do harm to nobody. Murder is a form of harm. Ergo: You must not commit murder.

'4 REINE RECHTSLEHRE § 34(j). St. Thomas Aquinas, SUMMA THEOLOGICA q. XCV, art. 2 (T. Gilby transl. 1966). (Aqui- nas' "second way" of deriving a command from a natural law is determination [determinatio]. See id.) Aquinas' "first way" provides an example of deducibility within a normative order. But as Aquinas' theory of "practical reason" makes abundantly clear, it would be a mistake to suppose that he is a "deductivist" who would subscribe to the tenets of Kelsen's static model. See id. at q. XCIV, art. 4. See also Ross, Justice is Reasonableness: Aquinas on Human Law and Morality, 58 MONIsT 86 (1974). 1975] Constraints on Legal Norms

Although Kelsen refers to the conclusion of Aquinas' deduction as a general legal norm, Kelsen's argument turns on the hypotheti- cal structure of that norm. In effect, Kelsen translates the norm into the equivalent legal proposition. For example, he interprets the command, "You must not commit murder" in terms of the legal proposition, "If you commit murder, then 0, a legal official, ought to impose on you a certain legal sanction." Kelsen's argument ex- ploits the idea that to instantiate the hypothetical, to "individual- ize" the general legal norm by applying it to a named person, in- volves an exercise of human will. This interposition of an act of will into what purports to be a purely logical operation severs Aquinas' deductive link at two points. The protasis of the hypothetical gives expression to what Kelsen terms "a specific condition," and the apodosis, to "a specific consequence,""6 and each marks a point at which Kelsen speaks of an act of individualization. Consideration of the protasis alone is sufficient to demonstrate the argument. The act of individualizing the protasis of the general norm "consists in establishing the actual existence ('in reality') of a cir- cumstance laid down in the general norm as the condition of a consequence. '57 Application of Aquinas' norm requires demonstrat- ing the satisfaction of the condition specified in the protasis, namely murder, by showing that, for example, a named individual Jones, acting with the requisite mental attitude, took the life of another person. But satisfaction of the protasis, Kelsen continues, is not something that can be established by the "purely logical function of subsuming an instance under a universal. 5 8 Whether acts of a certain class can be logically deduced or subsumed is irrelevant; what is critical, Kelsen contends, is that a particular act is puta- tively subsumed. Although a particular act a may not in reality be subsumable under the condition specified in the protasis of the general norm, if the appropriate legal official declares it subsuma- ble, then for legal purposes it is. If act a is in reality subsumable, but the legal official declares that it is not, then for legal purposes it is not. 9 Again, what is critical for Kelsen is not subsumability, but the act of putative subsumption. But if this argument proves anything about systemic substan- tive constraints, it proves too much. The very consideration that Kelsen brings to bear on this issue-that the actions of the legal

" The Idea of Natural Law, EssAYs at 41. ' Id. Id. at 42. , Id. at 47. The University of Chicago Law Review [42:768 official are decisive-applies with equal force to the doctrine of legal validity in the Pure Theory. That is, if the act of individuali- zation severs the deductive link in Aquinas' argument, it also severs the link between authorizing norm and individual norm in Kelsen's own doctrine of legal validity. One might argue that although a legal official's act may not be within the scope of the authorizing norm, if the official nevertheless declares that it is, then for legal purposes it is. So, the argument goes too far and, in fact, generates -the view, roughly, that the law is what officials say and do. And the problems with that view have been aired by, among others, Kelsen himself.6" Kelsen's effort to reject the systemic substantive constraints of the natural law theory comes as no surprise, for it is a central tenet of legal positivism that law may have any content whatever. Cu- riously, however, until very recently Kelsen accepted systemic con- straints on form, which he used in an attempt to resolve express conflicts between norms in the extraordinary case.

IV. CONSTRAINTS ON THE FORM OF LEGAL NORMS Kelsen suggests that express conflicts between norms are re- solved in the ordinary case by employing one or another of a variety of familiar nonsystemic or ad hoc rules and institutionalized prac- tices. For example, an express conflict between norms enacted at different times, either by the same legal organ or different legal organs having the same competence, may be resolved by invoking the rule lex posterior derogat priori.6 If the express conflict is be- tween simultaneously enacted norms, which precludes application of the rule lex posterior derogat priori, legal officials may have dis- cretionary power to choose one norm over the other, or they may resolve the conflict by interpretation. But sometimes no resolution of the conflict by recourse to non- systemic rules or practices may be forthcoming. Must the conflict stand if two norms, enacted at the same time, present an express conflict that legal officials are unable or unwilling to resolve by choice or interpretation? In the Pure Theory of Law (2d ed. 1960), Kelsen argues that in reality no conflict exists, for in a putative conflict one of the apparently conflicting norms is invalid.6 2 The

60 See, e.g., Kelsen, Eine "Realistische" und die Reine Rechtslehre [A "Realistic" Theory of Law and the Pure Theory of Law], 10 OSTERR, ZEITSCHR. F. 6FFENTL. REcHT 1 (1959). See also Hart, ScandinavianRealism, [1959] CAMB. L.J. 233. 11See REINE RECHTSLEHRE § 34(e). 12 See generally id. at §§ 16, 34(e); Kelsen, What is the Pure Theory of Law?, 34 TuL. L. 19751 Constraints on Legal Norms

argument turns on Kelsen's peculiar use of the principle of noncon- tradiction. The predicates "true" and "false" apply only to statements (or propositions). The principle of noncontradiction presupposes the applicability of these predicates and hence governs only statements (or propositions). Given two statements, in which the second ("Jones is not rich") denies exactly what the first asserts ("Jones is rich"), their conjunction is a self-contradiction-that is, the con- junction is always false. If the first statement is true, it implies the falsity of the second; and if the first statement is false, it implies the truth of the second. Applying the principle of noncontradiction directly to norms, in order to show that conflicting norms may stand in the same relationship to one another as conflicting statements, presupposes the applicability of the predicates "true" and "false" to norms. But as Kelsen himself points out repeatedly, a norm is neither true nor false but rather valid or invalid.63 If Kelsen were in fact attempting to apply the principle of non- contradiction directly to norms, the distinction between truth and validity would be enough to undermine his effort. But Kelsen is looking instead to an indirect or per analogiam application of the principle. When two norms appear expressly to conflict, a denial of legal validity to one of the norms is logically implied, in Kelsen's view, because the conjunction of certain statements corresponding to the norms is a self-contradiction. Kelsen writes: A norm is neither true nor false but rather, valid or invalid. What can be true or false, however, is a statement [Aussage], one that describes a normative order by asserting that accord- ing to this order a certain norm is valid-and, in particular, a legal proposition [Rechtssatz] that describes a legal order by asserting that according to this order a certain sanction ought or ought not, under certain conditions, to be imposed. Logical principles in general and the principle of noncontradictionin particular are applicable to legal propositions [Rechtssidtze] that describe legal norms and are, therefore, indirectly applica- ble to the legal norms themselves. It is thus not at all absurd to assert that two legal norms "contradict" one another. And therefore only one of the two can be regarded as objectively valid. 4

REV. 269, 271 (1960); Kelsen, Der Begriff der Rechtsordnung [The Concept of the Legal Order], 1 LOGIQUE Er ANALYSE 150, 155 (1958), in 2 DIE WIENER RECHTSTHEOREISCHE SCHULE 1395, 1401 (H. Klecatsky, et al., eds. 1968). ,3 See, e.g., Kelsen, What is the Pure Theory of Law?, 34 TuL. L. REV. 269, 271 (1960). '3 REINE RECHTSLEHRE § 34(e) (emphasis added). The University of Chicago Law Review [42:768 Kelsen begins his argument by considering the legal proposi- tions (Rechtssatze) that correspond to the expressly conflicting norms. Suppose, for example, that norm A is a command that Jones do act x at time t and that norm B is a prohibition, directing Jones to forbear from doing x to t. To all appearances the norms present an express conflict. The corresponding legal propositions, equiva- lent to norms A and B respectively, are: 1. If S, a legal subject, fails to do x at t, then 0, a legal official, ought to impose on S a certain legal sanction. 2. If S does x at t, then 0 ought to impose on S a certain legal sanction. Kelsen argues that where norms A and B appear to present an express conflict, one or the other is invalid because the conjunction of legal propositions 1 and 2 is a self-contradiction. But the argu- ment clearly will not do. A situation in which a legal subject is liable to a sanction whether or not he does x at t is anomalous to be sure, but an anomaly is not a self-contradiction. The conjunction of the legal propositions corresponding to expressly conflicting norms does not provide a reason for denying validity to one or the other of the norms. Although Kelsen applies his argument only to legal proposi- tions and not to the statements (Aussagen) used to assert that norms are valid or invalid, such statements present a parallel case for Kelsen's argument. Suppose, as before, that norm A is a com- mand that Jones do act x at time t and that norm B is a prohibition, bidding Jones to forbear from doing x at t. The statements (or, as expressed here, the statement-forms) used to assert that these norms are valid or invalid are: 1. Norm A is valid/invalid in this legal order. 2. Norm B is valid/invalid in this legal order. Thus, where norms A and B appear to present an express conflict, one or the other is invalid because the conjunction of statements 1. and 2. is a self-contradiction. Not surprisingly, the application of Kelsen's argument to state- ments fares no better than its application to propositions. Regard- less of how the predicates "valid" and "invalid" are assigned to norms A and B, the conjunction of the corresponding statements 1. and 2. is not self-contradictory. Statements 1. and 2. are used to assert or to deny that the norms exist (and the validity of a norm is its existence in a normative order). A conjunction of statements to the effect that one norm exists and a second does not is not self- contradictory; the conjunction is in every instance a contingent 1975] Constraints on Legal Norms statement that is, it may be either true or false. Although not demonstrating that the per analogiam argument for systemic constraints on form is mistaken, Kelsen retracts the argument in his recent papers on legal norms in the Essays and denies that such constraints are possible. He readily embraces the legal consequence of his new position, namely that in the case of expressly conflicting norms, both are valid. He writes: Conflicts between norms of morals and norms of law are famil- iar to everybody. But there are also conflicts of norms within one and the same legal order. . . as e.g. a conflict between two statutes or conflicts between norms of one and the same stat- ute. In all these cases both norms are valid, if there are no special legal provisions solving this conflict. Since both norms are valid, the one must be violated if the other is obeyed." There is no analogy of the required sort between validity and truth, and in the Essays Kelsen takes seriously the implications of this fact. That a norm is valid means that it is present. That a norm is not valid means that it is absent. An invalid norm is one that does not exist, and is thus not a norm. But a false statement is also a statement; it is present as a statement, even if it is false. 6 And, as a corollary, Kelsen notes that a norm has a temporal exist- ence, while the truth or falsity of a statement is non-temporal. A norm takes on validity, i.e., begins to be valid in time, and goes out of validity, i.e., ceases to be valid in time or loses its validity. A statement does not begin or cease to be true. If it is true, it always has been and always will be. 7 These and other aspects of the validity of legal norms were entirely familiar to Kelsen in other contexts, for example, in his illuminating discussions of the temporal, spatial (or territorial), personal, and material parameters of legal validity. 8 But only now, in the recent papers on legal norms in the Essays, has Kelsen brought the impli- cations of these aspects of validity to bear on his earlier efforts to understand express conflicts between norms in terms of the analogy of self-contradictory conjunctions of statements.

1SKelsen, Law and Logic, in PHILOSOPHY AND CHRISTIANITY, note 5 supra, at 233. 14 Law and Logic, ESSAYS at 230. 57 Id. 11See H. KELSEN, PRINCIPLES OF INTERNATIONAL LAW chs. 2-3 (1952). 786 The University of Chicago Law Review [42:768

CONCLUDING REMARK The aim of the Pure Theory is, as Kelsen puts it, "to free legal science from all extraneous elements."6 Having rejected the exist- ence of systemic constraints on the content of legal norms in many of his writings, and having now rejected systemic constraints on form, Kelsen has in the recent Essays expurgated from legal science the last extraneous elements of psychology, sociology, ethics, and political theory-"impurities" that stood in the way of an under- standing of law as a system of authorized acts of will. There is, to be sure, no nodding agreement with Kelsen's views. But consensus is not numbered among the desiderata of philosophical work. Origi- nality and logical acumen are, and these Kelsen has in abundance. We will continue to learn from him.

11REINE RECITrSLEHRE § 1.